The Nevada Bat
Technical Notes Archive

Measuring Nocturnal Light



Note: As of October, 2003 Clairex has discontinued the shipping and manfacturing of CDS components, including the component used in this technical note. I must leave it to the experimenter to test the possible applicability of this sensor design to other available CDS components.

  One of the environmental factors influencing the behavior of bats, and many other nocturnally active animals, is the amount of light present during various activities. I have casually noted that some bats, like M. yumanensis, seem to favor foraging over water in shadowed areas, and usually emerge later in the night. As a personal point of interest, I have wanted to add light intensity logging to my tool kit. After a bit of experimentation, I came up with a sensor configuration that actually works very well !!!  
  My first experiment simply employed the standard light sensing capabilities of a HOBO data logger.

As others before me have found, these work well for ambient light levels above dusk, and in buildings, but saturate in full daylight, and have floored before full darkness.

The standard HOBO sensor just doesn't provide the range of sensitivity needed for logging the low light levels associated with moonlight and moon shadowing.

  So then I decided to see if I could come up with an alternative design for a low light sensor. After several circuits were worked up and tested, using small silicon photodiodes, and various photoconductive cells, I determined that the most promising configuration employed a CdS cell, which could be powered by the logger. The resulting circuit does an excellent job of generating a stable analog voltage that is very sensitive in extremely low light levels. I dubbed it the MoonLight sensor. The logger in the image above, has the standard internal light sensor, and an external MoonLight sensor described here. This configuration was used to produce the following graph.  

  The graph above shows the comparative response of both the standard HOBO light intensity sensor, plotted in blue, and the MoonLight sensor, plotted in pink, over a 12 hour period centered about midnight. The standard light sensor is bottomed out during the entire evening. The MoonLight sensor comes out of saturation, and displays it's working range of sensitivity as the full moon rises and descends.  

  Above is another plot sequence made with the MoonLight sensor. This time data from 5 consecutive nights have been merged to display the relative shifts in time and light intensity as the moon phase passes from waxing, to full, and then waning. The night of the 28th shows a particularly variable response as the moon was repeatedly obscured by passing clouds. The long descender at the end of the plot for 2/25 was created by wrapping the logger and sensor in a thick black blanket, and putting it in a dark closet for 10 minutes. This essentially provided a reference point that represented "absolute" darkness.  
The schematic diagram of the
MoonLight sensor is shown to the left. Note that it is a very simple device to build, as long as you use the correct components. The sensor circuit was developed specifically to work with an Onset Hobo logger.

The circuit derives power from the logger via the
red wire from the logger voltage cable. The white wire is the sensor voltage tracked by the logger, and the black wire is the logger's reference ground.

The photo cell I was using is a Clairex part number CL705HL401. It has maximum spectral response at 550nm, and excellent variability in extremely low light levels.
The value of Rs in the circuit determines the sensitivity and scale of the sensor. A value of 10 Megohms provides characteristics comparable to the plots shown above. Values can be shifted to 1 Megohm, which modifies the scale to cover brighter light levels, but sacrifices the expanded dark level coverage.

NOTE: The value of Rs should always be greater than 5K to prevent damage to the HOBO logger.

The MoonLight sensor works pretty well, as you can see from the graphs .... but there is a caveat in using these sensors! I am informed that the industry standard for specification tolerances of CDS photocells is typically 30% .... this amounts to a very wide latitude of variability in photo-conductive cells. Higher tolerances are achieved in critical applications by performing special part screening at the manufacturer's level, followed by further circuit modifications that allow fine tweaking of the sensor's characteristics to calibrate the finished product. This is what would be commonly done if you were making a high end ( $$$$$ ) light meter.

With that in mind, you should understand that the
MoonLight sensors may vary somewhat in sensitivity from one to the next, and can be difficult to calibrate as a group to one specific standard.

In my application, I am monitoring nocturnal light over a long period of time. So I can establish the full moon "calibration" level for the sensor deployed for that specific measurement station. From that point, I can scale all of the data from that particular sensor to the full moon level.



I thought I'd close this page out with this helpful utility that
lets you check the current lunar phase sequence.


Special thanks to Bill Rainey, for prodding me on, and the loan of his light meter !


Tony Messina, Las Vegas, NV - page last updated 11/17/2003